NDT Advance Access originally published online on July 4, 2007
Nephrology Dialysis Transplantation 2007 22(12):3631-3637; doi:10.1093/ndt/gfm420
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Pneumonitis associated with sirolimus: clinical characteristics, risk factors and outcome—a single-centre experience and review of the literature
1Department of Nephrology, Marienhospital Herne, Hospital of the Ruhr-University of Bochum, Hölkeskampring 40, 44625 Herne, 2Department of Nephrology, Krankenhaus der Barmherzigen Brüder Trier, Academic Clinic of the University of Mainz, Nordallee 1, 54292 Trier and 3Chirurgische Klinik, Knappschaftskrankenhaus, Hospital of the Ruhr-University of Bochum, Germany
Correspondence and offprint requests to: Stefan M. Weiner, Department of Nephrology, Krankenhaus der Barmherzigen Brüder Trier, Nordallee 1, D-54292 Trier, Germany. Email: stefan.weiner{at}ruhr-uni-bochum.de
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. The introduction of sirolimus as an immunosuppressive drug for renal transplantation has lead to an increase of unexplained interstitial pneumonitis.
Methods. Out of a cohort of 115 patients receiving sirolimus for prophylaxis of renal and/or pancreas transplant rejection, 11 patients with interstitial pneumonitis were identified. Medical records and published case series were reviewed to identify risk factors associated with the occurrence of pneumonitis.
Results. Eleven out of 80 patients (14%) with late switch to sirolimus developed pneumonitis, in contrast to none of the 35 patients with de novo use of sirolimus. The mean sirolimus trough level at presentation was 16.7 µg/l (range: 6.2–38.7 µg/l). Glomerular filtration rate (GFR) was significantly lower in patients with pneumonitis compared to controls (mean 21.3 ± 3.9 ml/min vs 38.65 ± 2.14 ml/min P = 0.002). Two patients needed haemodialysis shortly before pneumonitis was diagnosed. In a multivariate analysis only serum creatinine and GFR were independent predictors for pneumonitis.
Sirolimus was discontinuated in five patients and the dose reduced in the other patients. Pneumonitis resolved within 14–28 days in all patients. One patient who had continued low-dose sirolimus treatment relapsed after 5 months, the other five patients had no relapse over a period of 15–48 months.
Pooled analysis of our data and other published case series showed that the frequency of pneumonitis in patients with de novo use of sirolimus is significantly lower than in patients with late switch [5/133 (4%) vs 46/326 (14%) patients, P = 0.0024].
Conclusions. Late switch to sirolimus and impaired renal function are risk factors for pneumonitis. A sirolimus blood trough level above 12 µg/l may increase the risk, but pneumonitis may also occur at blood trough levels as low as 6 µg/l. Since pneumonitis may recur during low-dose sirolimus treatment, discontinuation of sirolimus appears to be the safest treatment option.
Keywords: pancreas transplantation; pneumonitis; rapamycin; renal transplantation; risk factor; sirolimus
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The introduction of sirolimus (rapamycin) as an immunosuppressive drug for renal transplantation has lead to an increase of unexplained interstitial pneumonitis. A recent review reported 43 patients with sirolimus-induced pulmonary toxicity, observed after renal, liver, heart and heart-lung transplantation [1]. Diagnosis is made by exclusion of infective, autoimmune and toxic causes of lung injury. The pneumonitis responds to drug withdrawal, however, regression of pneumonitis is also described after dose reduction of sirolimus [2,3]. Factors that may be related to the risk for sirolimus-related pneumonitis may include late switch to sirolimus, age, underlying disease, concomitant immunosuppressive treatment and serum levels of sirolimus [3].
Here, we report 11 cases of suspected sirolimus-induced pneumonitis. The medical records of our patients and the literature were reviewed to assess the risk factors of this drug-induced pulmonary toxicity. Furthermore, we report our experience with dose reduction of sirolimus as a treatment option of pneumonitis in six patients.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Medical records of patients from a single transplantation centre receiving sirolimus for prophylaxis of renal and/or pancreas transplant rejection were reviewed. Overall 115 patients were treated with sirolimus in our centre, 35 patients received low-dose sirolimus 2 mg/day in combination with tacrolimus (n = 27) and mycophenolate mofetil (MMF; n = 6) as first line prophylaxis of renal transplant rejection. Eighty patients were switched to sirolimus due to chronic transplant rejection or renal calcineurin inhibitor toxicity. Thirty-one patients had a combination treatment with tacrolimus, four patients with ciclosporine A and nine patients with MMF. Eighty-one patients underwent kidney transplantation, 33 patients had a combined kidney and pancreas transplantation and one patient had an isolated pancreas transplantation.
All patients with pneumonitis underwent computed tomography scans of the lungs. Bronchoalveolar lavage (BAL) was performed to exclude infectious agents, except in two patients due to anti-coagulation with cumarine and refusal to give informed consent. Medical records were reviewed and sirolimus trough levels measured by immunoassay to identify risk factors associated with the occurrence of pneumonitis. Glomerular filtration rate (GFR) was calculated by the modification of diet in renal disease (MDRD) formula according KDOQI.
We also reviewed all reports with more than one case report of sirolimus-associated lung toxicity to discuss the risk factors, treatment and outcome of pneumonitis.
Statistical methods included the Mann–Whitney rank sum test, chi-squared test and binary logistic regression to evaluate possible predictors for pneumonitis. The following biological parameters have been tested in univariate analyses: age, gender, sirolimus trough level, serum creatinine, GFR, proteinuria, time since transplantation, de novo use or late switch to sirolimus, immunosuppressants other than sirolimus and kidney disease. Multivariate analysis was used to validate whether possible predictors are related to pneumonitis independently from other biological parameters. P-values less than 0.05 were considered statistically significant. We calculated the coefficients Exp (B) (equivalent-to-odds ratio) and their 95% confidence intervals (95% CI). All analyses were conducted with the use of SPSS (Chicago, IL, USA, Version 11.5.1).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Out of 115 patients receiving sirolimus, 11 patients (9.6%) with interstitial pneumonitis were indentified. The patient characteristics are summarized in Table 1. Nine patients received sirolimus in combination with low-dose prednisone (range 2–10 mg/day), one patient received a triple combination with sirolimus and tacrolimus and one patient combined with ciclosporine A and prednisone. None of the 35 patients with de novo use of sirolimus developed pneumonitis (P = 0.02 vs patients with pneumonitis, chi-squared test).
|
Clinical presentation
The patients presented with low-grade fever (n = 8), dry cough (n = 7), dyspnoea on exertion (n = 4), fatigue (n = 3), and weight loss (n = 2), haemoptysis (n = 1) and chills (n = 1). Fever or weight loss was the sole symptom at presentation in three patients, whereas in the remaining eight patients pulmonary symptoms lead to the diagnosis of pneumonitis. Leucocyte counts varied markedly (mean 7433/µl, range 1100–12 000/µl and C-reactive protein levels were elevated (mean 11.1 mg/dl, range 1.1–23.5 mg/dl) (Table 2).
|
Comparison of patients with and without pneumonitis
Male gender was equally distributed in both groups (64% of patients, respectively). The mean age was not significantly different in both groups [52.6 ± 3.47 years (SEM) vs 47.5 ± 1.33 years (SEM) in patients with or without pneumonitis], as well as the kidney disease leading to end-stage renal disease. The mean sirolimus trough level at diagnosis of pneumonitis was 16.7 µg/l (range: 6.2–38.7 µg/l) (Table 2). However, the mean sirolimus trough levels within the first year after institution of sirolimus did not differ in patients with or without pneumonitis (9.75 µg/l ± 3.57 vs 10.01 µg/l ± 3.83). Transplant renal function was severely impaired in all patients with pneumonitis (mean serum creatinine: 3.1 mg/dl, range: 1.9–5.3 mg/dl). Two patients needed haemodialysis shortly before pneumonitis was diagnosed. The mean GFR was significantly lower in patients with pneumonitis (21.3 ± 3.9 ml/min SEM) compared with patients without pneumonitis (38.65 ± 2.14 ml/min SEM), P = 0.002 (Mann–Whitney rank sum test) (Figure 1). Tacrolimus was significantly more frequently used in patients without pneumonitis (n = 57; 55%) vs patients with pneumonitis (n = 1; 9%) (P = 0.01). The other immunosuppressive drugs were not different in patients without and with pneumonitis: prednisone (n = 98; 94% vs n = 11; 100%), mycophenolate (n = 16; 15% vs n = 0), cyclosporine A (n = 3; 3% vs n = 1; 9%), azathioprine (n = 1; 1% vs n = 0), P = NS for all.
|
Univariate analysis (binary logistic regression) revealed serum creatinine [P = 0.007; Exp(B) (95% CI)=1.738 (1.165–2.593)] and low GFR (MDRD) [P = 0.011; Exp(B) (95% CI)=0.916 (0.857–0.980)] as significant predictors for pneumonitis. The risk of pneumonitis was significantly lower in patients using tacrolimus [P = 0.02; Exp(B) (95% CI)=0.084 (0.01–0.68)]. Other parameters (age, gender, proteinuria, sirolimus trough levels, kidney disease, other immunosuppressants, de novo or late use of sirolimus) were not predictive for pneumonitis. Multivariate analysis showed that serum creatinine [P = 0.014; Exp(B) (95% CI)=1.902 (1.141–3.170)] and GFR [P = 0.020; Exp(B) (95% CI)=0.909 (0.839–0.985)] were predictive for pneumonitis independent of the other biological parameters tested, whereas the effect of tacrolimus was dependent of de novo use of sirolimus.
Lung imaging and bronchoalveolar lavage
Chest radiography was without pathological findings in four patients. CT scans of the lungs revealed bilateral mostly peripheral interstitial infiltrates in seven patients, accompanied with ground-glass opacities in five patients. Four patients presented with ground-glass opacities only (Figure 2). BAL was performed in nine patients. Blood cultures and cultures obtained from bronchial aspirates were negative. Cells obtained from BAL were analysed in three patients, showing lymphocytic alveolitis in two patients (71% and 83% lymphocytes) with a CD4/CD8 ratio within normal range. One patient had alveolar haemorrhage in BAL analysis. In two out of eight patients, PCR detected sequences of Mycoplasma pneumoniae (patient 5), influenza virus A (patient 3) and cytomegalovirus (CMV) (patient 3). Mycoplasma serology, blood CMV antigen (PP65) and CMV DNA repeatedly tested negative. Antibiotic and anti-viral treatment did not lead to improvement of the condition in these two patients.
|
Treatment and outcome
Sirolimus was discontinuated in five patients and the dose of sirolimus reduced in the remaining six patients. Antibiotic treatment was given in 10 of the 11 patients without an obvious benefit. Pneumonitis resolved within 14–28 days in all patients. Patient 4 continued sirolimus at a dose of 4 mg/day and a sirolimus blood level of 10 µg/l, but relapsed after 5 months. At this time, the sirolimus-level was elevated again (15.8 µg/l). After discontinuation of sirolimus pneumonitis regressed within 2 weeks. The remaining five patients with continued sirolimus medication had no relapse in the follow-up period of a mean of 29 months (range 15–48 months). The sirolimus dose was reduced to 1 mg/day in three patients, 3 mg and 6 mg in one patient, respectively. The mean sirolimus-level 1 month after resolution of pneumonitis was 8.92 µg/l (range 7.9– 10.5 µg/l) and was held in this range during the follow-up period.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The frequency of interstitial pneumonitis appears to be increased in renal transplant patients receiving sirolimus. Our study shows a frequency of 11 cases with pneumonitis out of 115 patients exposed to this drug, consistent with the estimated frequency of 5–11% according to previous studies [3,4]. Table 3 summarizes the literature including case series with more than one reported case [1,3,5–10].
|
Pneumonitis developed within 1–8 months after initiation of sirolimus therapy in accordance with the literature with the onset of symptoms within 1 and 51 months [3].
The clinical presentation of sirolimus-induced pneumonitis is heterogenous; fever occurs in 73% of patients, dyspnoea and cough may be the leading symptoms in most patients [1–3]. In our case series, fever was present in eight out of 11 patients and the leading symptom in two patients. Thus, pneumonitis should be included in the differential diagnosis of patients receiving sirolimus who present with fever of unknown origin.
Chest radiography was without pathological findings in four patients, whereas chest CT showed ground-glass opacities in all but one patient with a different severity. Six patients had additional peripheral infiltrations with a bronchiolitis obliterans-organizing pneumonia pattern. Thus, chest CT is recommended in cases of suspected initial pneumonitis, since especially ground-glass opacities may not be detectable in conventional chest radiographs. Parenchymal fibrosis may be the result of ongoing pneumonitis in some cases [1], indicating that early diagnosis may be crucial for the prognosis of pneumonitis.
BAL was performed in nine out of 11 patients. Two patients were PCR positive for mycoplasma, influenza or CMV. However, this does not prove mycoplasma pneumonitis. False-positive results are not a rare event [11] and mycoplasma serology was negative in our patients. The patient with positive CMV and influenza virus PCR was negative for pp65 antigen test as well as for blood CMV DNA. Furthermore, antibiotic and anti-viral treatment was without effect on the symptoms of the patients. In contrast, discontinuation of sirolimus lead to complete regression of pneumonitis.
BAL cellular analysis was performed in three patients, two of them showed lymphocytic alveolitis, a typical finding in most cases of sirolimus-induced peumonitis [2]. The third patient had alveolar haemorrhage, a rare phenomenon also described in a previous study [2]. In contrast, recent reports estimated a frequency of alveolar haemorrhage in 8–20% of patients with suspected sirolimus-associated pneumonitis [1–3].
Biopsy of the lung was not performed in our patients. However, several case reports showed that sirolimus-associated pulmonary toxicity is histologically characterized by the presence of organizing pneumonia (BOOP), interstitial pneumonitis, focal fibrosis or by the presence of alveolar haemorrhage. Since sirolimus-induced pneumonitis is not rare and the rapid improvement after discontinuation of sirolimus is helpful in the diagnosis, lung pathology may not be a routine diagnostic procedure of choice due to its high invasivity and costs.
Late switch from calcineurin inhibitors to sirolimus may increase the risk: although univariate analyses of our cohort have failed to find a significant effect of late switch on the risk of pneumonitis, an effect is evident when combining our data with previous studies showing a significantly lower frequency of pneumonitis in patients with de novo use of sirolimus than in patients with late switch (5/133 vs 46/326 patients, P = 0.0024) (Table 3). The reason for the higher frequency of pneumonitis in patients with late switch to sirolimus is unclear; two possible factors may be involved: sirolimus dose and renal function.
In human experience, most individuals with pneumonitis had sirolimus trough concentrations ranging between 15 and 30 ng/ml [2,12,13]. These reports contrast with other case series of lower sirolimus trough concentrations: 8.3–11.2 ng/ml in three patients [5], in non-toxic levels (10–15 ng/ml) in further three patients [1] and a single patient with sirolimus trough levels between 6 and 10 ng/ml [14]. In our study, only six of 11 patients had a trough concentration above 15 ng/ml and two patients below 10 ng/ml, suggesting that low sirolimus trough levels do not exclude this diagnosis. However, our observations show that none of the 35 patients with low-dose sirolimus 2 mg/day developed pneumonitis suggest that low sirolimus trough levels may reduce the frequency of pneumonitis in renal transplant recipients.
In our study, all patients with pneumonitis had severely impaired renal function. Moreover, univariate analyses showed that serum creatinine and GFR are predictive factors for pneumonitis in our cohort. Furthermore, only serum creatinine and GFR are independent predictors for pneumonitis in multivariate analyses. Unfortunately, in several reports, the creatinine levels of the patients suffering pneumonitis are not given [2,3,6,9,10]. However, the three patients reported by Pham et al. [1] had elevated creatinine levels (1.9 mg/dl, 4.0 mg/dl and 2.9 mg/dl, respectively) as well. In another series of three cases, creatinine level was given in one patient (160 µmol/l) and a second patient was switched to sirolimus due to a progressive rise of plasma creatinine [5]. A moderately impaired renal function was also present in four patients in a recent report [8]. It is very likely that patients in other reports also had impaired renal function, since the main reason to switch to sirolimus may be advanced calcineurin inhibitor renal toxicity. The tolerance against sirolimus may be altered in the presence of severe renal insufficiency as it is well known for other drug toxicities e.g. neurotoxicity of calcineurin inhibitors in the presence of liver failure [15].
Our data do not support further risk factors for pneumonitis, such as age, gender, underlying disease or concomitant immunosuppressive agents. Univariate analysis showed a lower risk of pneumonitis in patients using tacrolimus, however, the result of the multivariate analysis suggests that this effect is rather a consequence of the lower frequency of pneumonitis in patients with de novo use of sirolimus than a hint for a protective value of tacrolimus. Thus, review of the literature showed that 10% of the patients suffering pneumonitis also used tacrolimus.
Treatment of choice of suspected sirolimus-induced pneumonitis is withdrawal of sirolimus. However, some authors also suggested dose reduction of sirolimus as treatment option. In our case series, sirolimus was discontinued in five patients and the sirolimus dose was reduced in the remaining six patients. One of the patients with continued sirolimus treatment had recurrent pneumonitis, however, both episodes were associated with a high sirolimus blood concentration. Recently, recurrence of pneumonitis during continued but reduced sirolimus dose was also reported by Champion et al. [3]. Here, we report the second case with recurrence of pneumonitis after continued low-dose sirolimus exposure, thus sirolimus blood concentrations should be monitored closely and patients should be informed about the possibility of recurrence of pneumonitis.
Steroid treatment of pneumonitis was also used in several case reports (Table 3), however, no convincing data are available to support this strategy. The favourable outcome in our patients without high-dose steroids suggests that steroid treatment is not required in this setting. In accordance with our experience, most patients recovered without sequelae also in other case series. Rarely, development of pulmonary fibrosis and death has been reported [5,7], thus, pneumonitis should be recognized and treated early to avoid a fatal outcome.
The precise aetiology of sirolimus-induced pneumonitis is unknown. Elevated eosinophils and mast cells as described by Morelon et al. [2] suggest hypersensitivity against sirolimus as cause of pneumonitis. Our observations that dose reduction of sirolimus leads to regression of pneumonitis favours direct drug toxicity at least in some patients.
Conflict of interest statement. Independent from this study, Prof. Rump received a research grant from Wyeth Pharmaceuticals. No other potential conflict of interest relevant to this article was reported.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Pham PT, Pham PC, Danovitch GM, et al. Sirolimus-associated pulmonary toxicity. Transplantation (2004) 77:1215–1220.[CrossRef][Web of Science][Medline]
- Morelon E, Stern M, Israel-Biet D, et al. Characteristics of sirolimus-associated interstitial pneumonitis in renal transplant patients. Transplantation (2001) 72:787–790.[CrossRef][Web of Science][Medline]
- Champion L, Stern M, Israel-Biet D, et al. Brief communication: sirolimus-associated pneumonitis: 24 cases in renal transplant recipients. Ann Intern Med (2006) 144:505–509.
[Abstract/Free Full Text] - Fairbanks KD, Eustace JA, Fine D, Thuluvath PJ. Renal function improves in liver transplant recipients when switched from a calcineurin inhibitor to sirolimus. Liver Transpl (2003) 9:1079–1085.[CrossRef][Web of Science][Medline]
- Haydar AA, Denton M, West A, Rees J, Goldsmith DJ. Sirolimus-induced pneumonitis: three cases and a review of the literature. Am J Transplant (2004) 4:137–139.[CrossRef][Web of Science][Medline]
- Dominguez J, Mahalati K, Kiberd B, McAlister VC, MacDonald AS. Conversion to rapamycin immunosuppression in renal transplant recipients: report of an initial experience. Transplantation (2000) 70:1244–1247.[CrossRef][Web of Science][Medline]
- Lindenfeld JA, Simon SF, Zamora MR, et al. BOOP is common in cardiac transplant recipients switched from a calcineurin inhibitor to sirolimus. Am J Transplant (2005) 5:1392–1396.[CrossRef][Web of Science][Medline]
- Morath C, Schwenger V, Ksoll-Rudek D, et al. Four cases of sirolimus-associated interstitial pneumonitis: identification of risk factors. Transplant Proc (2007) 39:99–102.[CrossRef][Web of Science][Medline]
- Delgado JF, Torres J, Jose Ruiz-Cano M, et al. Sirolimus-associated interstitial pneumonitis in 3 heart transplant recipients. J Heart Lung Transplant (2006) 25:1171–1174.[CrossRef][Web of Science][Medline]
- Chhajed PN, Dickenmann M, Bubendorf L, Mayr M, Steiger J, Tamm M. Patterns of pulmonary complications associated with sirolimus. Respiration (2006) 73:367–374.[CrossRef][Web of Science][Medline]
- Ursi D, Ieven M, Noordhoek GT, Ritzler M, Zandleven H, Altwegg M. An interlaboratory comparison for the detection of Mycoplasma pneumoniae in respiratory samples by the polymerase chain reaction. J Microbiol Methods (2003) 53:289–294.[CrossRef][Web of Science][Medline]
- Singer SJ, Tiernan R, Sullivan EJ. Interstitial pneumonitis associated with sirolimus therapy in renal-transplant recipients. N Engl J Med (2000) 343:1815–1816.
[Free Full Text] - McWilliams TJ, Levvey BJ, Russell PA, Milne DG, Snell GI. Interstitial pneumonitis associated with sirolimus: a dilemma for lung transplantation. J Heart Lung Transplant (2003) 22:210–213.[CrossRef][Web of Science][Medline]
- Digon BJ 3rd, Rother KI, Hirshberg B, Harlan DM. Sirolimus-induced interstitial pneumonitis in an islet transplant recipient. Diabetes Care (2003) 26:3191.
[Free Full Text] - Bechstein WO. Neurotoxicity of calcineurin inhibitors: impact and clinical management. Transpl Int (2000) 13:313–326.[CrossRef][Web of Science][Medline]
Accepted in revised form: 4. 6.07
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Alexandru, A. Ortiz, S. Baldovi, J. M. Milicua, E. Ruiz-Escribano, J. Egido, and J. J. Plaza Severe everolimus-associated pneumonitis in a renal transplant recipient Nephrol. Dial. Transplant., October 1, 2008; 23(10): 3353 - 3355. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||